Research on stochastic resonance enhancement of X-ray images based on a genetic algorithm

摘要

Due to the poor enhancement of a faint image in X-ray images, a novel method based on the scale-variable stochastic resonance model is proposed in this paper. Firstly, the effect of the stochastic resonance model parameters on the enhanced image quality is analysed. Then, for obtaining the optimum model parameters, a stochastic resonance image enhancement method based on a genetic optimisation algorithm is proposed. The genetic algorithm employs an image quality evaluation index as the fitness function, which can indicate the variety of image quality. Making full use of parallel optimisation of the genetic optimisation algorithm and weak signal enhancement of the stochastic resonance, the proposed method overcomes the poor enhancement of traditional methods. The proposed method was applied to simulated images and typical defect images. Experimental results confirmed its effectiveness in increasing the visibility of images.